AGENT SKILL

video-detection-workflows

Operate ImageAI video and camera object detection workflows with callbacks for standard and custom models.

License 许可
MIT
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
video-detection-workflows
license
MIT
metadata
{ "disco-role": "operating" }
description
Operate ImageAI video and camera object detection workflows with callbacks for standard and custom models.
disable-model-invocation
true

02 资源文件 5

03 SKILL.md 正文

Video Detection Workflows

Use this sub-skill when the task is to run, script, debug, or explain ImageAI object detection over video files, explicit OpenCV camera/live-stream inputs, or custom ImageAI-trained video detection models.

Route Here For

  • Standard COCO video detection with VideoObjectDetection and RetinaNet, YOLOv3, or TinyYOLOv3 weights.
  • Camera or live-stream detection through an explicitly created cv2.VideoCapture object.
  • Standard-detector class filtering with CustomObjects for selected COCO classes.
  • Custom video detection with CustomVideoObjectDetection, .pt/.pth weights, and the matching JSON detection config.
  • Save/no-save operation, output path behavior, FPS, frame detection intervals, timeouts, progress logging, display flags, and video analytics callbacks.

Route Elsewhere

  • Still-image object detection, extraction, image arrays, or COCO class details belong to object-detection-workflows.
  • Custom model training, dataset layout, Pascal VOC conversion, or interpreting newly trained artifacts belong to custom-training-and-data.
  • Install/import/backend setup and model-asset acquisition policy belong to the root image-ai skill.

Operating Checklist

  1. Choose detector family:
    • Standard COCO: VideoObjectDetection; model type retinanet, yolov3, or tiny-yolov3.
    • Custom trained detector: CustomVideoObjectDetection; model type yolov3 or tiny-yolov3, plus matching JSON config.
  2. Choose exactly one input source. Prefer input_file_path for files. Open a camera only when the user explicitly requests a camera index or stream URL.
  3. Decide saving. If save_detected_video=True, provide an output base path; ImageAI appends .mp4 and returns that path string. If save_detected_video=False, use callbacks for useful output; the detection call returns None.
  4. For cameras and long files, set detection_timeout, choose a practical frames_per_second, and raise frame_detection_interval on slow hardware.
  5. Match callback arity to return_detected_frame. The complete-video callback never receives a frame argument.
  6. Use the bundled helper when a safe command-line wrapper is enough: scripts/detect_video.py.

References

04 同仓库更多 Skills

VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-triton-opt

Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-solve

END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-skill

PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-pattern-fusion

Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-orchestrate

Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-feedback

Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).

★ 351